Temperature measurement by infrared thermography is a technique that is widely used in predictive maintenance to detect faults. The uncertainty involved in measuring temperature by thermography is not only due to the imager, but also due to the measurements and estimates made by the user: emissivity of the inspected object, distance, temperature, and relative humidity of the propagation medium, temperature of objects located in the ambient, and the imager itself. This measurement uncertainty should be available for the thermographer to be able to make a more accurate diagnosis. The methods available in the literature to estimate the uncertainty of measured temperature usually require information nonaccessible to the regular thermographer. This paper proposes a method for calculating the uncertainty of temperature that requires only data available to the thermographer. This method is useful under usual conditions in predictive maintenance—short distance (7.5 to 14 μm) thermal imagers, no fog or rain, among others. It provides results similar to methods that use models that are not available or reserved by the manufacturers of imagers. The results indicate that not all sources of uncertainty are relevant in measurement uncertainty. However, the total uncertainty can be so high that it may lead to misdiagnosis.